A complete diallel cross, including inbreds and reciprocals, was made among six inbred lines of mice. Body weight, tail length and tail width were measured at ages of 1 through 12 weeks. The analysis described by Griffing (1956a, 1956b, 1958) as the modified diallel, method 3, model II was made for each trait at each age, a separate analysis being made for each sex. Inbreds did not contribute to estimates of the effects in the diallel model, but were used to estimate heterosis. Positive heterosis was observed for all three traits. Heritability increased with age for all three traits, although it remained small for tail width; for body weight it was larger in females than for males, while for tail length and width the opposite was true. Non-additive genetic variance was observed for all three traits. Maternal effects variance was virtually non-existent for tail length, but for body weight and tail width exhibited a marked peak around weaning, followed by a gradual decline for body weight and a rapid decline for tail width. Environmental variance exhibited a marked peak at weaning for all three traits and was larger for male body weight and tail length from four weeks onward. Residual reciprocal effects were important for tail length at all ages, but were small or negligible for body weight and tail width. It is concluded that the relative importance of the individual's genotype in determining size increases with age, while that of non-genetic factors declines.